postgres/typesense
1
0
Fork 0
mirror of https://github.com/typesense/typesense.git synced 2025-05-18 20:52:50 +08:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

Fix max_candidates behavior.

Browse Source
This commit is contained in:
Kishore Nallan 2022-05-13 12:45:21 +05:30
parent d83745a6eb
commit 0e2257dc29
4 changed files with 63 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
include/index.h
View File

@ -583,6 +583,7 @@ private:
const std::vector<bool>& prefixes,
bool prioritize_exact_match,
const bool exhaustive_search,
const size_t max_candidates,
int syn_orig_num_tokens,
const int* sort_order,
std::array<spp::sparse_hash_map<uint32_t, int64_t>*, 3>& field_values,

2
src/collection_manager.cpp
View File

@ -866,7 +866,7 @@ Option<bool> CollectionManager::do_search(std::map<std::string, std::string>& re
}
if(!max_candidates) {
max_candidates = exhaustive_search ? 10000 : 4;
max_candidates = exhaustive_search ? Index::COMBINATION_MAX_LIMIT : Index::COMBINATION_MIN_LIMIT;
}
Option<nlohmann::json> result_op = collection->search(raw_query, search_fields, simple_filter_query, facet_fields,

7
src/index.cpp
View File

@ -1156,6 +1156,7 @@ void Index::search_all_candidates(const size_t num_search_fields,
const std::vector<bool>& prefixes,
bool prioritize_exact_match,
const bool exhaustive_search,
const size_t max_candidates,
int syn_orig_num_tokens,
const int* sort_order,
std::array<spp::sparse_hash_map<uint32_t, int64_t>*, 3>& field_values,
@ -1166,7 +1167,7 @@ void Index::search_all_candidates(const size_t num_search_fields,
auto product = []( long long a, tok_candidates & b ) { return a*b.candidates.size(); };
long long int N = std::accumulate(token_candidates_vec.begin(), token_candidates_vec.end(), 1LL, product);
long long combination_limit = exhaustive_search ? Index::COMBINATION_MAX_LIMIT : Index::COMBINATION_MIN_LIMIT;
long long combination_limit = std::max<size_t>(Index::COMBINATION_MIN_LIMIT, max_candidates);
for(long long n = 0; n < N && n < combination_limit; ++n) {
RETURN_CIRCUIT_BREAKER
@ -2647,7 +2648,7 @@ void Index::fuzzy_search_fields(const std::vector<search_field_t>& the_fields,
long long n = 0;
long long int N = std::accumulate(token_to_costs.begin(), token_to_costs.end(), 1LL, product);
const long long combination_limit = exhaustive_search ? Index::COMBINATION_MAX_LIMIT : Index::COMBINATION_MIN_LIMIT;
const long long combination_limit = std::max<size_t>(Index::COMBINATION_MIN_LIMIT, max_candidates);
while(n < N && n < combination_limit) {
RETURN_CIRCUIT_BREAKER
@ -2764,7 +2765,7 @@ void Index::fuzzy_search_fields(const std::vector<search_field_t>& the_fields,
sort_fields, token_candidates_vec, searched_queries, qtoken_set, topster,
groups_processed, all_result_ids, all_result_ids_len,
typo_tokens_threshold, group_limit, group_by_fields, query_tokens,
num_typos, prefixes, prioritize_exact_match, exhaustive_search,
num_typos, prefixes, prioritize_exact_match, exhaustive_search, max_candidates,
syn_orig_num_tokens, sort_order, field_values, geopoint_indices,
query_hashes, id_buff);

57
test/collection_specific_more_test.cpp Normal file
View File

@ -0,0 +1,57 @@
#include <gtest/gtest.h>
#include <string>
#include <vector>
#include <fstream>
#include <algorithm>
#include <collection_manager.h>
#include "collection.h"
class CollectionSpecificMoreTest : public ::testing::Test {
protected:
Store *store;
CollectionManager & collectionManager = CollectionManager::get_instance();
std::atomic<bool> quit = false;
std::vector<std::string> query_fields;
std::vector<sort_by> sort_fields;
void setupCollection() {
std::string state_dir_path = "/tmp/typesense_test/collection_specific_more";
LOG(INFO) << "Truncating and creating: " << state_dir_path;
system(("rm -rf "+state_dir_path+" && mkdir -p "+state_dir_path).c_str());
store = new Store(state_dir_path);
collectionManager.init(store, 1.0, "auth_key", quit);
collectionManager.load(8, 1000);
}
virtual void SetUp() {
setupCollection();
}
virtual void TearDown() {
collectionManager.dispose();
delete store;
}
};
TEST_F(CollectionSpecificMoreTest, MaxCandidatesShouldBeRespected) {
std::vector<field> fields = {field("company", field_types::STRING, true)};
Collection* coll1 = collectionManager.create_collection("coll1", 1, fields).get();
for (size_t i = 0; i < 200; i++) {
nlohmann::json doc;
doc["id"] = std::to_string(i);
doc["company"] = "prefix"+std::to_string(i);
ASSERT_TRUE(coll1->add(doc.dump()).ok());
}
auto results = coll1->search("prefix", {"company"}, "", {}, {}, {0}, 10, 1, FREQUENCY, {true}, 0,
spp::sparse_hash_set<std::string>(),
spp::sparse_hash_set<std::string>(), 10, "", 30, 4, "title", 20, {}, {}, {}, 0,
"<mark>", "</mark>", {}, 1000, true, false, true, "", false, 6000 * 1000, 4, 7,
false, 1000).get();
ASSERT_EQ(200, results["found"].get<size_t>());
collectionManager.drop_collection("coll1");
}